In the manufacturing industry, different paradigms and manufacturing models impose different rules on the manufacture of a product. One of these manufacturing models is called “lean manufacturing” or “lean inventory management.” The lean manufacturing model attempts to eliminate waste in the manufacturing process by creating a stable production environment. One way of creating stability involves imposing a leveling constraint in the production environment. Production of a product may occur during a period of time, which may be divided into multiple windows or “buckets” of time. A leveling constraint dictates that the same amount of the product should be produced during each time window.
The lean manufacturing model attempts to remain responsive to customer needs, supplying customers with products when the customers need the products, while eliminating waste in the production environment. One way manufacturers attempt to maintain responsiveness is to store some quantity of a product in inventory as “safety stock.” Because manufacturers typically cannot predict customer demand for a product with absolute certainty, forecasts or predictions of customer demand typically include some amount of error. The forecast error means that customer demand for the product may be higher or lower than predicted. Safety stock helps protect the manufacturer when customer demand for the product is higher than expected. When the manufacturer cannot produce enough of the product to satisfy the higher customer demand, the manufacturer may use the safety stock to meet the customer demand.
A problem with this approach is that safety stocks typically fail to protect a manufacturer from both forecast error in the demand forecast and customer demand spikes. A customer demand spike refers to a period when demand for a product is greater than surrounding time periods. In some lean manufacturing environments, the goal is to produce the same amount of a product during each window of time. When a customer demand spike occurs during later time windows, the manufacturer may produce enough of the product during the earlier time windows to meet this higher demand. When a customer demand spike occurs during earlier time windows, the manufacturer may be unable to produce enough of the product to meet this early demand. While safety stocks may protect a manufacturer against early customer demand for a product, it typically cannot protect the manufacturer both from forecast error in the demand forecast and from early customer demand spikes. This leaves the manufacturer vulnerable to “stock-outs,” or times when the manufacturer cannot meet customer demand because no product is in stock.
Another problem with this approach is that identifying a quantity of a product to protect against both forecast error in the demand forecast and customer demand spikes is difficult. It is often a simple task to identify a quantity of a product to use as a safety stock. A manufacturer typically uses the average and standard deviation of customer demand for a product to determine an appropriate size for the safety stock. It is typically more difficult to identify a quantity of a product to protect against customer demand spikes in a lean manufacturing environment. The amount of inventory needed to protect against customer demand spikes is typically unrelated to the forecast error of the demand forecast. As a result, manufacturers often attempt to guess how much inventory is needed to protect against customer demand spikes, which typically results in inaccurate estimates. When demands associated with multiple customers are considered, it becomes even more difficult to predict the quantity of the product needed to protect against customer demand spikes, which results in even more inaccurate estimates.
As a result of any of these or other disadvantages, previous lean inventory management techniques have been inadequate in many manufacturing environments.